A realização do sujeito pronominal no português popular de Salvador

Abstract

Many scholars have researched different aspects of Rizzi's Null-Subject Parameter proposal, with the result that inadequacies in the original proposal have led to newer formulations. The current sociolinguistic study follows aspects of the original formulation of the Null Subject Parameter hypothesis as studied by a number of researchers of Brazilian Portuguese, and investigates the sociolinguistic profile of the pronominal subject in a corpus of urban Brazilian Portuguese drawn from suburbs of Salvador, in the state of Bahia. The study was conducted within the framework of the Vertentes Project on Bahian Vernacular Portuguese and follows the assumptions of Variationist Sociolinguistics. The corpus analyzed comprises 48 interviews, distributed between the suburbs of Cajazeiras, Liberdade, Plataforma and Itapuã. The dependent variable studies is the form (realized or null) of the subject pronoun. Eight independent linguistic variables are evaluated: person, verbal inflection, previous reference to the subject, semantic characterization of the subject, type of sentence, localization of the subject pronoun in relation to its antecedent, topicalization of the subject and presence of elements to the left of the verb. The effects of four social variables are evaluated: gender, age group, schooling, periods spent outside the community, and locality. The data were submitted to the GoldVarbX program package for quantitative processing. The results demonstrate a greater pronominal performance of 57.8% versus 42.2 unrealized subject. The results identify as conditioning factors of the pronominal subject: person, type of sentence, reference previous to subject, position in relation to antecedent, presence of constituent to the left of the verb, and age group. Thus, the study contributes to the discussions about the pro drop parameter and to the understanding of the Brazilian linguistic reality regarding an aspect of its grammar.